A PROPOSED RADIOGRAPHIC CLASSIFICATION SCHEME FOR

CONGENITAL THORACIC VERTEBRAL MALFORMATIONS IN

BRACHYCEPHALIC “SCREW-TAILED” DOG BREEDS

RODRIGO GUTIERREZ-QUINTANA, JULIEN GUEVAR, CATHERINE STALIN, KITERIE FALLER,

CARMEN YEAMANS, JACQUES PENDERIS

Congenital vertebral malformations are common in brachycephalic “screw-tailed” dog breeds such as French
bulldogs, English bulldogs, Boston terriers, and pugs. The aim of this retrospective study was to determine
whether a radiographic classification scheme developed for use in humans would be feasible for use in these
dog breeds. Inclusion criteria were hospital admission between September 2009 and April 2013, neurologic
examination findings available, diagnostic quality lateral and ventro-dorsal digital radiographs of the tho-
racic vertebral column, and at least one congenital vertebral malformation. Radiographs were retrieved and
interpreted by two observers who were unaware of neurologic status. Vertebral malformations were classified
based on a classification scheme modified from a previous human study and a consensus of both observers.
Twenty-eight dogs met inclusion criteria (12 with neurologic deficits, 16 with no neurologic deficits). Congeni-
tal vertebral malformations affected 85/362 (23.5%) of thoracic vertebrae. Vertebral body formation defects
were the most common (butterfly vertebrae 6.6%, ventral wedge-shaped vertebrae 5.5%, dorsal hemivertebrae
0.8%, and dorso-lateral hemivertebrae 0.5%). No lateral hemivertebrae or lateral wedge-shaped vertebrae
were identified. The T7 vertebra was the most commonly affected (11/28 dogs), followed by T8 (8/28 dogs)
and T12 (8/28 dogs). The number and type of vertebral malformations differed between groups (P = 0.01).
Based on MRI, dorsal, and dorso-lateral hemivertebrae were the cause of spinal cord compression in 5/12
(41.6%) of dogs with neurologic deficits. Findings indicated that a modified human radiographic classification
system of vertebral malformations is feasible for use in future studies of brachycephalic “screw-tailed” dogs.
C 2014 American College of Veterinary Radiology.

Key words: butterfly vertebra, congenital vertebral malformation, hemivertebra, kyphosis, scoliosis.

Introduction breeds is due to hemivertebra affecting the coccygeal re-

H EMIVERTEBRAE, WEDGE-SHAPED (or cuneiform) verte-
brae, and butterfly vertebrae are relatively common
congenital vertebral malformations in dogs, and are asso-
ciated with varying degrees of kyphosis and scoliosis.1,2 Al-
though any breed can be affected, they appear to be more
common in brachycephalic “screw-tailed” breeds such as
the English bulldog, French bulldog, Boston terrier, and
Pug.1–5 It has been suggested that the kinked tail in these
From the School of Veterinary Medicine, College of Medical, Veteri-
nary and Life Sciences, University of Glasgow, Bearsden Road, Glasgow,
G61 1QH, UK.
An abstract with limited preliminary results of this study was sub-
mitted to the European College of Veterinary Neurology Annual Sym-
posium, September 2013, Paris, France.
Address correspondence and reprint requests to Rodrigo Gutierrez-
Quintana. School of Veterinary Medicine, College of Medical, Veteri-
nary and Life Sciences, University of Glasgow, Bearsden Road, Glasgow,
G61 1QH, UK. E-mail: Rodrigo.GutierrezQuintana@Glasgow.ac.uk.
(Rodrigo Gutierrez-Quintana)
Received December 22, 2013; accepted for publication February 19,
2014.
doi: 10.1111/vru.12172
gion and this trait has been selected as a desirable pheno-
type for many generations.3,4,6 These congenital vertebral
malformations can occur in isolation or be multiple in an
individual dog, and they occur more frequently in the tho-
racic vertebral column.2–5 In general, these malformations
are considered to be an incidental radiographic finding be-
cause many affected dogs do not show any neurological
signs.1,2,4,5 When clinical signs do occur they are consid-
ered to be a consequence of progressive vertebral canal
stenosis and vertebral column instability. Vertebral lesions
may be apparent at birth, but clinical signs mostly only de-
velop later in life.1,2,7 The etiology of congenital vertebral
malformations remains unclear, but may be due to congen-
ital absence of vertebral vascularization, genetic defects or
teratogenic insult to active cartilaginous proliferation.1,2,4
In humans, congenital vertebral malformations have
been classified into categories on the basis of their radio-
graphic appearance.9–11 The first classification category in-
cludes failure of vertebral segmentation, in which portions
Vet Radiol Ultrasound, Vol. 00, No. 0, 2014, pp 1–7.

1
2 GUTIERREZ-QUINTANA ET AL.
TABLE 1. Clinical 0 to 5 Points Grading Scale for Thoracolumbar Spinal
Cord Lesions13
Grade Clinical signs
0 Normal
1 Spinal pain
2 Ambulatory paraparesis
3 Nonambulatory paraparesis
4 Paraplegia with intact deep pain perception
5 Paraplegia with absent deep pain perception
of adjacent vertebral elements fail to divide (block vertebra
or bars). The second category consists of failure of verte-
bral formation, in which a portion of a vertebral element
is deficient (hemivertebra, wedge shape vertebra or but-
terfly vertebra). Based on these categories, Nasca et al.10
proposed a radiographic classification for congenital sco-
liosis and McMaster et al.11 for congenital kyphosis and
kyphoscoliosis. One previous canine study used the Nasca
classification system in five clinical cases and one canine
abstract classified wedging of the vertebral body as uni-
lateral, dorsal, and ventral.5,12 To the authors’ knowledge,
no study has evaluated the McMaster radiographic clas-
sification scheme for use in brachycephalic “screw-tailed”
dogs.
In this retrospective study, our primary objectives were
(1) to develop and apply a radiographic classification
scheme for congenital vertebral malformations affecting
the thoracic vertebral column in a group of brachycephalic
“screw-tailed” dogs and (2) to test associations between the
type of vertebral malformation and the presence of neuro-
logical deficits.
Materials and Methods
The medical records of the University of Glasgow
Small Animal Hospital were retrospectively reviewed from
September 2009 to April 2013 to identify French bulldogs,
English bulldogs, Boston terriers, and Pugs with or without
neurological deficits, that had well positioned lateral and
ventro-dorsal digital radiographs of the thoracic vertebral
column, with at least a single vertebral congenital malfor-
mation present. The age, sex, and breed were recorded. If
any neurological deficits were associated with the vertebral
malformation, then the duration of clinical signs and the
neurological grade at presentation were recorded. The neu-
rological grade was determined using the standard 0 to 5
grading system (Table 1)13 . Patients were then divided into
two groups, one with neurological deficits localized to the
thoracic spinal cord segments (Group 1) and one without
neurological deficits (Group 2).
Radiographs for each dog were retrieved and evaluated
by two observers who were unaware of dog group status
at the time of interpretation. Images were displayed using
an open-source PACS Workstation DICOM viewer (Osirix
2014
Imaging Software, v 3.9.2, Pixmeo, Geneva, Switzerland).
Each vertebral malformation classification was determined
by a consensus between the two observers. The number and
the vertebral level were also recorded for each vertebral
malformation and for each patient. When available, MRI
studies were also retrieved and interpreted.
The vertebral malformations were classified as defects of
segmentation if adjacent vertebral elements fail to divide
(block vertebrae) or defects of formation if a portion of the
vertebra was deficient. Defects of formation were then sub-
classified into ventral aplasia of the vertebral body (dorsal
hemivertebra), lateral aplasia of the vertebral body (lateral
hemivertebra), ventro-lateral aplasia of the vertebral body
(dorso-lateral hemivertebra), ventral and median aplasia of
the vertebral body (butterfly vertebra), ventral hypoplasia
of the vertebral body (ventral wedge shape vertebra), and
lateral hypoplasia of the vertebral body (lateral wedge shape
vertebra) using a modified classification based on that of
McMaster et al.11 (Fig. 1). The presence of any other verte-
bral malformations, including thoracic and cervical transi-
tional vertebrae, spina bifida and any other types, were also
recorded.
Statistical tests were selected by a statistician. Data were
analyzed using commercially available software (Minitab
16 Statistical Software, Minitab Inc. State College, PA,
USA). Descriptive statistics were reported as mean, range,
and standard deviation (SD). A Fisher exact test was used
to compare the vertebral malformations between the two
groups with significance defined as P < 0.05. For that the
number of dogs with severe vertebral malformations (ven-
tral aplasia of the vertebral body, lateral aplasia of the verte-
bral body, and ventro-lateral aplasia of the vertebral body)
and with less severe vertebral malformations (ventral and
median aplasia of the vertebral body, ventral hypoplasia of
the vertebral body, lateral hypoplasia of the vertebral body,
and other unclassified vertebral malformations) in each of
the groups was calculated.
Results
Twenty-eight dogs were included in the study. Eleven
Pugs, 11 English bulldogs, three French bulldogs, and
three Boston terriers. The mean age was 2.5 years (range:
4 months - 14 years, SD: 2.86 years). There were 20 males
(two neutered) and eight females (one neutered). Twelve
dogs had neurological deficits (Group 1) and 16 dogs did
not have neurological deficits (Group 2) at the time of
thoracic spinal radiography. The 12 dogs with neurological
deficits (Group 1) included nine Pugs, two English bulldogs
and one French bulldog. The mean age was 2.4 years
(range: 4 months - 7.5 years, SD: 2.4 years). There were six
males (one neutered) and six females (one neutered). Nine
dogs presented with ambulatory paraparesis and ataxia
VOL. 00, NO. 0 CLASSIFICATION OF VERTEBRAL MALFORMATIONS 3

FIG. 1. Schematic line drawing showing the radiographic appearance of the vertebral malformation classification system used in this study.

(grade 2), two with nonambulatory paraparesis (grade 3)
and one with paraplegia (grade 4). The mean duration
of neurological deficits prior to presentation was 68 days
(range: 1–240 days, SD: 69 days). All Group 1 dogs
had MRI of the thoracic spine using the same 1.5 Tesla
scanner (Magnetom Essenza, Siemens, Camberley, United
Kingdom), and spinal cord compression was identified in
all of them. Sagittal T2-weighted images and transverse
T2-weighted and T1-weighted images of the area of
interest were acquired in all these cases. In 10 of the dogs,
MRI confirmed that spinal cord compression was directly
caused by the vertebral malformation with no other
4 GUTIERREZ-QUINTANA ET AL. 2014

FIG. 2. Congenital vertebral malformations identified on lateral radiographs of the thoracic vertebral column. (A) Block vertebrae, where there is failure
of normal segmentation of the vertebrae, evident as the failure of division of T8 and T9 vertebral bodies (open arrow) and spinous process (solid arrow). (B)
Ventral hypoplasia of T8 vertebral body (ventral wedge shape vertebra) (arrow) with kyphosis. (C) Ventral aplasia of T8 vertebral body (dorsal hemivertebra)
(arrow) with more severe kyphosis.

FIG. 3. Congenital vertebral malformations identified on ventro-dorsal radiographs of the thoracic vertebral column. (A) Ventral hypoplasia of T8 vertebral
body (arrow, this is the same case as figure 2B): no evidence of scoliosis is evident on the ventro-dorsal view if there is no lateral hypoplasia. (B) Ventro-lateral
aplasia of T8 vertebral body (dorso-lateral hemivertebra) (arrow). This case had similar changes to figure 2C on the lateral view. (C) Ventral and median aplasia
of the vertebral body (Butterfly vertebra) (arrow).

findings. One dog had an intervertebral disc herniation
adjacent to the vertebral malformation and another dog
had discospondylitis and secondary empyema in the
intervertebral disc adjacent to the vertebral malformation.
The 16 dogs with no neurological deficits (Group 2)
included nine English bulldogs, three Boston terriers, two
Pugs, and two French bulldogs. The mean age was 2.6
years (range: 7 months - 14 years, SD: 3.2 years). There
were 14 males (one neutered) and two females. None of
the Group 2 dogs had MRI of the thoracic spine.
In all dogs, radiography of the thoracic vertebral column
was performed using the same digital radiography system
(Siemens, Camberley, United Kingdom). A total of 362
vertebrae were evaluated, as two of the dogs had 12 tho-
racic vertebrae. A congenital malformation was identified
in 85 (23.5%) vertebrae, with 18 (64.2%) of the dogs hav-
ing multiple malformations. It was possible to classify 50
of these vertebral malformations using the proposed clas-
sification (Fig. 1). Of the 35 vertebral malformations that
failed to meet the criteria for classification, 26 vertebrae
VOL. 00, NO. 0 CLASSIFICATION OF VERTEBRAL MALFORMATIONS 5

TABLE 2. Number of Vertebrae Affected by Each Type of Congenital neurological deficits (Table 2). Butterfly vertebrae were not
Vertebral Malformation in Dogs with Associated Neurological Deficits
(Group 1) and Dogs Without Associated Neurological Deficits
found to be the cause of the spinal cord compression in
(Group 2). A Significant Difference in the Types of Congenital Vertebral MRI studies of any of the 10 dogs where the spinal cord
Malformations was Evident when Comparing the Two Groups compression was directly caused by the congenital vertebral
(P = 0.01)
malformation.
Vertebral malformation All dogs Group 1 Group 2
Block vertebra 1 1 0
Ventral aplasia of the 3 3 0
vertebral body (Dorsal Discussion
hemivertebra)
Ventro-lateral aplasia of the 2 2 0 In the present study we proposed and applied a radio-
vertebral body graphic classification of congenital vertebral malforma-
(Dorso-lateral
hemivertebra)
tions in dogs based on previous human studies of congenital
Lateral aplasia of the 0 0 0 kyphotic, scoliotic, and kypho-scoliotic malformations.10,11
vertebral body (Lateral This classification scheme was helpful for more consistently
hemivertebra)
Ventral and median aplasia of 24 4 20
describing the different types of vertebral malformations
the vertebral body present in these breeds, specifically the defects of verte-
(Butterfly vertebra) bral body formation. Similar to humans, defects of verte-
Ventral hypoplasia of the 20 6 14
vertebral body (Ventral
bral formation were more common in the dogs included
wedge shape) in this study than defects of vertebral segmentation.11 The
Lateral hypoplasia of the 0 0 0 results of the present study indicated that less severe ver-
vertebral body (Lateral
wedge shape)
tebral body formation defects such as ventral and median
Total number of vertebrae 50 16 34 aplasia (butterfly vertebra) or ventral hypoplasia of the ver-
with a classified congenital tebral body (ventral wedge shape vertebra) are common
malformation
Total number of vertebrae 35 14 21
and can be seen in dogs with and without neurological
with an unclassified deficits. Where spinal cord compression was directly caused
congenital malformation by the congenital vertebral malformation, none were due
Total number of vertebrae 362 154 208
evaluated
to ventral and median aplasia of the vertebral body (but-
terfly vertebra) in this sample population. More severe de-
fects of vertebral body formation, including ventral and
ventro-lateral aplasia of the vertebral body (dorsal and
were judged as having a shorter vertebral body length with dorso-lateral hemivertebrae), were more likely to be asso-
a normal shape, four had fused spinous processes with nor- ciated with neurological deficits. These findings are similar
mal vertebral bodies, three were transitional vertebrae and to human patients, where the severity of defects of verte-
two demonstrated spina bifida. Of the 50 vertebral malfor- bral body formation has been shown to be proportional
mations classified, one was a defect of vertebral segmenta- to the severity of the kyphosis.11 Another interesting find-
tion and 49 were defects of vertebral formation. The most ing of the current study was that no lateral aplasia or hy-
common malformation was butterfly vertebra seen in 24 poplasia of the vertebral body was observed. It is possible
vertebrae (6.6%), followed by ventral wedge shape vertebra that kyphosis and kypho-scoliosis are more frequent and
in 20 vertebrae (5.5%), dorsal hemivertebra in three ver- important than scoliosis in brachycephalic “screw-tailed”
tebrae (0.8%), dorso-lateral hemivertebra in two vertebrae dog breeds. It is also important to notice that of the 35
(0.5%) and block vertebra in one vertebra (0.3%) (Figs. 2 vertebral malformations that failed to meet the classifi-
and 3). No lateral hemivertebra or lateral wedge shape ver- cation criteria, 26 were judged as having a shorter verte-
tebra were identified in this study (Table 2). The most com- bral body length with a normal shape. For future studies,
monly malformed vertebra was T7 (11 dogs), followed by the authors recommend that this type of defect of verte-
T8 (eight dogs) and T12 (eight dogs). Figure 4 demonstrates bral body formation should be added to the classification
the distribution of the type of malformation by vertebral scheme as symmetrical hypoplasia of the vertebral body
column level. (“short” vertebra) (Fig. 5).
There was a significant difference in the number of verte- Findings in the current study were consistent with pre-
bral malformations between groups (P = 0.01). Dorsal and vious studies indicating that congenital vertebral malfor-
dorso-lateral hemivertebrae were only present in Group 1 mations are common in the thoracic vertebral column of
dogs, and these were interpreted to be the cause of neu- brachycephalic “screw-tailed” dog breeds, with 23.5% of
rological deficits in 5/12 (41.6%) of these dogs. Butterfly the vertebrae examined in this study being affected. The
and wedge shape vertebrae were identified in both groups, majority of the dogs (64.2%) in this study had multi-
but were more commonly seen in Group 2 dogs with no ple vertebral malformations. This finding was also similar
6 GUTIERREZ-QUINTANA ET AL. 2014

FIG. 4. Histogram showing the frequency of defects of vertebral body formation by vertebral column level.

to previous reports describing a high frequency of verte-

bral abnormalities in French bulldogs, English bulldogs,
and other brachycephalic breeds with multiple vertebrae
commonly affected.2–5 Also similar to previous studies was
our observation that the most commonly affected vertebra
was in the midthoracic region (T6-T9 region).1,2,5
The neurological signs were compatible with a T3-L3
spinal cord segment myelopathy in all the cases with neu-
rological deficits. The majority of dogs with neurological
signs (75%) presented with a chronic and progressive his-
tory of ambulatory paraparesis and ataxia (Grade 2), which
suggested slowly progressive spinal cord injury and/or
compression, consistent with previous reports.1,2,7,14 Pugs FIG. 5. Schematic line drawing showing the radiographic appearance of
were noted to have neurological deficits in 75% of the the additional vertebral malformation found frequently in this study: sym-
patient population evaluated here. Two possibilities for the metrical hypoplasia of the vertebral body (“short” vertebra).
over-representation of neurological deficits in this breed
may be that there are more severe vertebral malformations
present in this breed, or that they were more common in our There were several limitations to this study, including it
population. In two of the 12 cases with neurological deficits, being retrospective in nature. Second, the breed distribu-
it was found that the clinical signs were not a direct conse- tion between the two groups was not similar and this could
quence of vertebral canal stenosis caused by the vertebral have influenced the type of vertebral malformations seen
malformation, but were secondary to another neurological in each group. Third, the number of cases in the present
disease that occurred concurrently. In both cases the inter- study was small, thus it was not possible to give an accurate
vertebral discs adjacent to the vertebral malformation were representation of the frequency of the different vertebral
the ones involved (one dog with Hansen type I interverte- malformations in each breed. Another limitation was the
bral disc herniation and the other dog with discospondylitis absence of long-term follow up of the cases and it is un-
and secondary empyema). It is possible that intervertebral known if some of the dogs in Group 2 may have developed
instability between the normal and malformed vertebrae neurological deficits at a later date. None of them had any
could have resulted in damage to the intervertebral disc additional imaging to confirm that there was no spinal cord
and predisposed it to herniation or infection. compression at the site of their vertebral body anomalies.
VOL. 00, NO. 0 CLASSIFICATION OF VERTEBRAL MALFORMATIONS
The authors feel that this is an important consideration
when interpreting the results of this study.
In conclusion, radiographic classification of congenital
vertebral malformations in brachycephalic “screw-tailed”
dogs is feasible. Defects of vertebral body formation were
the most frequent in this sample population. More severe
defects of vertebral body formation, such as ventral and
ventro-lateral aplasia (dorsal and dorso-lateral hemiver-
tebrae), may more likely be associated with neurological
deficits, but further studies in a larger group of dogs are
needed to confirm this.
REFERENCES
1. Westworth DR, Sturges BK. Congenital spinal malformations in
small animals. Vet Clin North Am Small Anim Pract 2010;40:951–
981.
2. Moissonier P, Gossot P, Scotti S. Thoracic kyphosis associated with
hemivertebra. Vet Surg 2011;40:1029–1032.
3. Done SH, Drew RA, Robins GM, Lane JG. Hemivertebra in the dog:
clinical and pathological observations. Vet Record 1975;96:313–317.
4. Schlensker E, Dislt O. Prevalence, grading and genetics of hemiverte-
brae in dogs. Eur J Comp Anim Pract 2013;23:119–123.
5. Volta A, Morgan JP, Gnudi G, Bonazzi M, Gazzola M, Zanichelli
S et al. Clinical-radiological study of the vertebral abnormalities in the En-
glish bulldog. Proceeding of the 12th annual conference of the European
Association of Veterinary Diagnostic Imaging; 2005 Oct 5–8; Naples, Italy:
p. 31.
6. Kramer JW, Schiffer SP, Rantanen NW Whitener EK. Characteriza-
tion of heritable thoracic hemivertebra of the German shorthaired pointer. J
Am Vet Med Assoc 1982;181:814–815.
7. Aikawa T, Kanazono S, Yoshigae Y, Sharp NJH, Muñana KR. Ver-
tebral stabilization using positively threaded profile pins and polymethyl-
7
Conflict of Interest Statement
None of the authors has any financial or personal rela-
tionships that could inappropriately influence or bias the
content of the paper.
ACKNOWLEDGMENTS
The authors thank Dr. Timothy Parkin, Boyd Orr Centre for Population
and Ecosystem Health, School of Veterinary Medicine, College of Med-
ical, Veterinary and Life Sciences, University of Glasgow for assistance
with statistical analyses.
methacrylate, with or without laminectomy, for spinal canal stenosis and
vertebral instability caused by congenital thoracic vertebral anomalies.Vet
Surg 2007;36:432–441.
8. Pourquie O, Kusumi K. When body segmentation goes wrong. Clinical
Genetics 2001;60:409–416.
9. Jaskwhich D, Ali RM, Patel TC, Green DW. Congenital scoliosis.
Curr Opin Pediatr 2000;12:61–66.
10. Nasca RJ, Stelling FH, Steel HH. Progression of congenital scolio-
sis due to hemivertebrae and hemivertebrae with bars. J Bone Joint Surg
1975;57(A):456–466.
11. McMaster MJ, Singh H. Natural history of congenital kyphosis and
kyphoscoliosis. J Bone Joint Surg 1999;81(A):1367–1383.
12. Besati O, Ozak A, Pekcan Z, Eminaga S. Nasca classification of
hemivertebra in five dogs. Ir Vet J 2005:58:688–690.
13. Sharp NJ, Wheeler SJ. Thoracolumbar disc disease. In: Sharp NJ,
Wheeler SJ. (eds.), Small animal spinal disorders: Diagnosis and surgery.
2005;Elsevier Mosby, Philadelphia, USA: 121–159.
14. Jeffery ND, Smith P, Talbot CE. Imaging findings and surgical treat-
ment of hemivertebrae in three dogs. J Am Vet Med Assoc 2007;230:532–536.